This article has been accepted for publication in a future issue of this journal, but has not been

fully edited. Content may change prior to final publication. Citation information: DOI
10.1109/ACCESS.2020.3007205, IEEE Access

Date of publication xxxx 00, 0000, date of current version xxxx 00, 0000.
Digital Object Identifier 10.1109/ACCESS.2020Doi Number

Face Recognition Attendance System Based

on Real-time Video Processing
1, a 2, b*
Hao Yang and Xiaofeng Han
1
College of Media Engineering, Communication University of Zhejiang, Hangzhou 310018, Zhejiang, China
2
College of Mathematics and Systems Science, Shandong University of Science and Technology, Qingdao 266590, Shandong, China
Hao Yang (aEmail: 415360889@qq.com)
*
Corresponding author: Xiaofeng Han (bEmail: taianhaiyue@126.com)
This work was supported by Basic Public Welfare Research Project of Zhejiang Province (LGF20H180001) and Basic Public Welfare Research Project
of Zhejiang Province (LGF20F020001).

ABSTRACT With the advent of the era of big data in the world and the commercial value of face
recognition technology, the prospects for face recognition technology are very bright and have great
market demand. This article aims to design a face recognition attendance system based on real-time video
processing. This article mainly sets four directions to consider the problems: the accuracy rate of the face
recognition system in the actual check-in, the stability of the face recognition attendance system with
real-time video processing, the truancy rate of the face recognition attendance system with real-time video
processing and the interface settings of the face recognition attendance system using real-time video
processing. By analyzing the situation of these problems, the concept of attendance system based on face
recognition technology is proposed, and the research on face recognition attendance system based on
real-time video processing is carried out. Experimental data shows that the accuracy rate of the video face
recognition system is up to 82%. Compared with the traditional check-in method, the face recognition
attendance system can be reduced by about 60%. The rate of skipping classes has greatly reduced the
phenomenon of students leaving early and skipping classes. The face recognition time and attendance
system with real-time video processing through the above experimental certification can quickly complete
the tasks of students in the time and attendance check-in system, get rid of the complex naming
phenomenon, greatly improve the efficiency of class, and play an important role in guiding the
development of the time and attendance system.

KEY WORDS Video Processing, Face Recognition Technology, Face Recognition Attendance,
Attendance System, Video Recognition

I. INTRODUCTION world, especially today, when terrorist activities are rampant,

In this era of Internet explosion, computer technology has this technology has received more and more attention. Face
involved many areas of people's lives and work. The recognition technology has many typical applications in the
occasions where people come into contact with computers field of public safety, civil economy, and home entertainment
are gradually expanding. The frequency with which people [1-2]. The pipeline of general enterprises needs to record the
use computing is also increasing. One of the most attendance of personnel, which has become a basic
challenging projects in the field has a broad application requirement of the company. However, when these
prospect because of its huge sense of innovation. As an attendance systems are formulated, unnecessary errors often
important identity label for people to distinguish different occur. Taking the current fingerprint attendance system as an
individuals, face recognition technology has gradually example, the study has found that The fingerprint attendance
entered people's lives. Face recognition is the combination system has an error rate of about 5%, and there will be a
of artificial intelligence and computer. Because of its huge phenomenon that fingerprints cannot be hit, which seriously
challenging innovation and broad application prospects, it affects the efficiency of attendance, especially in large
has become the most challenging topic in this field. attendance sites, which is more likely to cause congestion.
In recent years, the face recognition application system has However, the card attendance system has the phenomenon of
developed rapidly as a computer security technology in the
VOLUME XX, 2020 1

This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/.
 This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI
10.1109/ACCESS.2020.3007205, IEEE Access
Author Name: Face Recognition Attendance System Based on Real-time Video Processing

employees swiping cards for someone else, and it is difficult quickly completed the tasks, got rid of the complicated sign
to achieve the purpose of real time attendance. Compared of roll call, and soon realized the sign of operation and
with the two attendance systems, the face recognition system function. The future system time and the form of attendance
has higher accuracy and stability, because there are more system conversion have made tremendous innovations,
points for face recognition, which is more accurate than other greatly improving the attendance rate and the reliability of
systems. Greatly improved, it is difficult to congestion [3-4]. face recognition technology. It is worthy of further
Although China's research on face recognition technology exploration and realization by our scientists.
started late, our scientific researchers have caught up and
some leading figures have established their own industry II. PROPOSED METHOD
positions in the field of face recognition. With the advent of
the era of big data in today's world and the commercial value A. REAL-TIME VIDEO FACE IMAGE RECOGNITION
of face recognition technology, the prospect of this (1)Face recognition
technology research is very bright and has great market Face recognition is the core of the entire recognition
demand [5]. process. Face recognition is a computer vision technology
Faces in surveillance videos often suffer from serious that analyzes facial feature information for identity
image blur, posture changes, and occlusion. In order to identification. In a broad sense, face recognition is divided
overcome the challenges of video-based face recognition into two parts: face detection and face recognition matching.
(VFR), Ding C proposed a comprehensive framework based Face recognition technology is based on the facial features of
on convolutional neural network (CNN). First, in order to the person, and the input face image or video stream. First
learn a fuzzy and robust face representation, Ding C determine whether there is a human face, if there is a human
artificially blurs the training data composed of clear still face, then further give the position, size of each face and the
images to make up for the lack of real video training data. position information of each major facial organ. Based on
Using training data composed of still images and artificial this information, the identity features contained in each face
fuzzy data, CNN is encouraged to automatically learn fuzzy are further extracted and compared with known faces to
insensitive features. Second, in order to enhance the identify the identity of each face [8-9]. Face recognition
robustness of CNN features to pose changes and occlusion, technology belongs to biometric recognition technology,
CNN has proposed a trunk branch CNN model (TBE-CNN), which mainly includes four parts: face image collection, face
which extracts complementarity from the overall face image image preprocessing, face image feature extraction, matching
and the patches around the face parts Information [6]. and combining hard recognition, combined with hardware
Scholars such as Nemirovskiy V B have studied the cameras, network lines and computing device. The
featureless face recognition problem. The recognition is calculation method is as follows:
based on clustering the proximity between the cardinal
distributions of the luminance clusters of the divided images. T = min{T1 , T2 , , Tn } (1)
As a proximity measure, Nemirovskiy V B uses three types
of distances: Euclidean distance, cosine distance, and Leibler The principle of face recognition. Face recognition
distance. Recursive neural network software model is used technology is a kind of biometrics technology, through the
for image segmentation and proximity measure clustering [7]. acquisition of camera equipment face information and
The design of face recognition attendance system with pre-processing. Face detection. The main purpose of face
real-time video processing is conducive to the development detection is to collect information to determine whether there
of enterprises and has a positive effect on the development of is a human face image in the image, and to determine the size
enterprises in the future. This article aims to design a face and position of the image, and segment the detected human
recognition time and attendance system based on real-time face image into the adult face area. The last link is face
video processing. In this experiment, four investigation recognition, extracting facial feature information and image
experiments were carried out. Among them, the accuracy rate information to determine whether it is in the repository. If it
of the face recognition system in actual check-in; the stability is, it has matching identity information, otherwise there will
of the face recognition time and attendance system with be no recognition results [10-12].
real-time video processing; analysis of the skip rate of face (2) Face feature extraction by LDA method
recognition attendance system using real-time video The Chinese meaning of LDA is linear discriminant
processing; interface settings of face recognition attendance analysis. This algorithm is to find a set of linear
system using real-time video processing. The experimental transformations that minimize the intra-class dispersion
results prove that the time and attendance system achieves the between each category and maximize the inter-class
expected time and attendance results through face recognition dispersion. Fisher is a linear judgment method in LDA. The
technology and with the help of a computer, which fully name is called because Fisher function is often used for
reflects the feasibility design of the overall algorithm. The calculation. LDA can also be used very well in face
students who completed the attendance sign-in system recognition, but when using this method to extract face

2 VOLUME XX, 2020

This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/.
 This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI
10.1109/ACCESS.2020.3007205, IEEE Access
Author Name: Face Recognition Attendance System Based on Real-time Video Processing

features, some small sample problems often occur. As long as The analysis method of subspace is to use spatial
a training sample of the face can be given, and the sample transformation to map face image data into a certain subspace
belongs to the other four categories, the total number of faces to achieve dimensionality reduction of face data. It is a huge
in the sample is: face data that becomes simple to calculate, and then to the
dimensionality reduction data classification. Different
M subspace analysis methods use different criteria, and different
N = ∑M1 (2) subspaces can be obtained. Common subspace analysis
t =1 methods for face recognition include: principal component
analysis, linear discriminant analysis, independent element
The average value among these categories is calculated as analysis, etc.
follows: 3) Neural network method
Neural network is a commonly used method in membrane
M recognition. Its principle is to use a large number of simple
1
Z=
M
∑Z
t =1
t
(3) calculation units to form a certain hierarchical structure. Each
simple unit can only solve simple calculations, but the system
composed of units in complex structures can be a
The dispersion within the sample is denoted as Sw, and the complicated question. The neural network algorithm has also
dispersion between the samples is denoted as Sb. The achieved good results in face recognition. For example,
calculation formula is: commonly used BP networks, self-organizing networks,
convolutional networks, etc. As long as the network is large
1 t enough, there are enough training samples to theoretically
t − z) Φ Φ
Sb = M t ( zt − z) ( z =
t
B B
(4) recognize all faces. Although neural networks have some
N
advantages in face recognition, they also have considerable
defects. The structure of neural networks is huge and
2
complex, and their training requires a huge sample library.
Φ = (M t N ) ( Z − z )
B
t (5) The training time often takes days or even months. The speed
is not fast enough. Therefore, neural networks are not
commonly used in the actual application of face recognition.
Φb = [Φ b1 , Φb 2 ,... Φbn] (6) 4) Support Vector Machine (SVM) method
Support vector machine is a research hotspot of pattern
recognition. The basic principle of the algorithm is to use
M M
1
∑∑ (Z
samples to form a lattice in the high-level feature space,
Sn = N t =1 t =1
M
− ZT) (7) select sample points as the support vector near the boundary
between the two types of sample points, and use the support
vector to make the decision. And finally achieve the purpose
(3) Main face recognition methods of classification and identification. The method of projecting
1) Geometric feature method into a high-dimensional space makes it possible to solve
Since the facial features such as eyes, nose, ears, mouth, many problems that are difficult to linearly classify in the
etc. are different in structure, different human faces are status space. But support vector machines also have their
represented according to different characteristics of the shortcomings. First, support vector machines are a two-class
characteristic shapes of these organs. Geometric features classification algorithm. Although some methods can be used
were first used in the description and recognition of the side to solve multi-classification problems, their efficiency is
profile of a human face. It determines a number of feature often greatly reduced. Secondly, support vector machines
points based on the profile line of the person's side, and then want high-dimensional space projection, which requires the
derives a set of feature quantities for recognition such as support of kernel functions, but choosing kernel functions is
angle and distance based on these feature points. Its indeed a lot of trouble. Finally, although support vector
advantage is the use of simple geometric information, so the machines can classify independently, the effect of directly
time cost of storage space and classification is small, and it classifying faces is not good, and often feature extraction is
can still be used when the image recognition rate is low; it is required for faces.
not sensitive to changes in lighting. Its disadvantage is that it The dimensionality reduction process is performed before
is difficult to extract stable features from the image, it is recognition, and the basic flow is shown in Figure 1 [14].
greatly affected by changes in posture and expression, and
the stability is not high [13].
2) Subspace analysis method

VOLUME XX, 2020 3

This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/.
 This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI
10.1109/ACCESS.2020.3007205, IEEE Access
Author Name: Face Recognition Attendance System Based on Real-time Video Processing

Main face recognition methods ignore it; if it has not been checked in, add the check in
information to the database check in table. The module will
return check-in information, including: whether the check-in
Geometric feature
Subspace analysis
Neural network Support Vector was successful; the student's name; the student's student ID;
method method Machine method
the course that was checked in; if the check-in was
unsuccessful, what is the reason.
FIGURE1. Main face recognition methods
4) The background management module is responsible for
the background administrator, which mainly includes the
functions of counting sign-in information, entering or
(4) Video image recognition system
deleting student information, etc.
The face detection technology is used to locate and
segment a partial face image from the image; the feature
B. BASIC FACE RECOGNITION ALGORITHM
extraction technology extracts the amount of data that can When designing a system to select a face recognition
characterize the face image, and forms the features to be algorithm, we have to consider the following factors:
stored in the feature database. Face recognition process: face recognition rate, algorithm robustness, and matching time
positioning and image processing in the image; feature [17-19]. To sum up, the system selects Gabor features plus
extraction and selection; detecting and recognizing the image, Fisher based discriminant analysis method based on
and returning the recognition result. First create a facial orthogonal basis to become a linear discrimination method
image file of the face. That is, use the camera to collect facial [20-21]. Face image feature representation is a key part in
image files of the person's face or take their photos to form a face recognition, and good feature representation can
facial image file, and store these facial image files to generate improve the robustness of image matching. Gabor wavelet
faceprint codes. Get the current human face. That is, use the feature description method is a comprehensive method that
current facial image captured by the camera, or take a photo combines gray and local descriptions. It has the advantages
input, and generate a facial texture code from the current of gray-based and feature-based methods [22].
facial image file. Compare with the current facial texture Gabor kernel definition of wavelet transform:
encoding and file inventory. That is, to retrieve and compare
the current facial texture code with the facial texture code in   
 k j2 k j2 x 2  σ2
the file inventory. The above-mentioned "face coding" Ψ j ( x ) = 2 exp(− )[exp(i k x ) − exp( − )] (8)
σ 2σ 2
j
method works according to the essential characteristics of the 2
human face and the beginning. This facial coding can resist
 
changes in light, skin tone, facial hair, hairstyle, glasses, x = ( x, y ) is a spatial domain variable, and k j is a
expression and posture, and has strong reliability, so that it
frequency vector that determines the scale and direction of
can accurately identify a person from millions of people. The 
face recognition process can be completed automatically, the Gabor kernel. We selected an image I (x ) , and the
continuously, and in real time using ordinary image 
x
Gabor transform at the 0 point at a specific position in
processing equipment [15]. The video image recognition
system is mainly composed of four parts: login module, the picture can be realized by convolution with Gabor
recognition module, check-in module and background kernel at this point. The face image is I (z ) , and the
management module [16]. Taking the school system as an z = ( x, y ) picture scale direction is µ ,ν , there is a
example, the functions of each module are as follows:
1) The login module is where the lecturer or background formula:
administrator logs in with an account and password to view
attendance information. Gµ ,ν ( z ) = I ( z ) ∗ψ µ ,ν ( z ) (9)
2) The main function of the recognition module is to
receive a face picture, call the system application With z = ( x, y ) , the calculation formula can be
programming interface (Application Programming Interface,
obtained:
API) to perform face recognition on the picture, and obtain
an identification code that uniquely determines the picture.
µ 2 +ν 2 ( µ 2 +ν 2 )( x 2 + y 2 ) σ 2 (10)
3) The check-in module receives the identification code G ( x, y ) = exp(− )[exp(i ( µx +νy )) − exp(− )]
σ 2
2σ 2
2
obtained in the identification module and compares it with
the student information in the database to find and confirm Among them, 。
µ k µax πM k πM π
the student information; by querying the current time and the k = ( ), µ = coσ( ),υ = µax σiν( ), k µax = , f = 2 , σ = 2π
ν fN 8 fN 8 2
schedule information in the database, the current course Fisher discriminant analysis is an improved algorithm
information is obtained. After obtaining the student and based on PCA. It adopts the method of dimensionality
course information, it is logical to determine whether the reduction to effectively reduce the amount of calculation,
student has checked in for this class, if it has been checked in,
4 VOLUME XX, 2020

This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/.
 This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI
10.1109/ACCESS.2020.3007205, IEEE Access
Author Name: Face Recognition Attendance System Based on Real-time Video Processing

thereby improving the calculation speed [23]. This method III. EXPERIMENTS
can ensure that the projected samples have the smallest
intra-class distance and the largest inter-class distance in the A. EXPERIMENTAL SETUP
new space, that is, the best separability in the space. (1)Experimental background
Introduce Fisher discriminant criteria table: In order to deeply study the application of face recognition
attendance system for real-time video processing, from the
accuracy of actual recognition of face recognition system, the
ϕ T Sbϕ
I fisher (ϕ ) = (11) stability of background application system of face recognition
ϕ T S wϕ attendance system, the face recognition system faces some
problems; the development status of the role played by the
Where ϕ is any n -dimensional column vector. face recognition system in the actual check-in, the factors
Fisher's linear discrimination method selects the vector restricting the development, and the difficulties faced by the
development are considered, and many problems are found.
ϕ that maximizes I fisher (ϕ ) as the projection direction, so (2)Experiment setup process
that the projected samples have the largest inter-class The experiment sets up a control group and an
dispersion and the smallest intra-class dispersion. Among experimental group, the control group uses traditional
them, S b is the inter-class separation matrix, and S w is the fingerprint check-in, and the experimental group uses a
real-time video processing face recognition attendance
intra-class separation matrix. Let Fk , k = 1,2, , L be
0
system. Choose two universities from a province, choose the
the average of the training samples of class ϕ k after PCA same number of students for the experiment, collect, count
and analyze the experiment, analyze the application space,
and LDA transformation. Use the nearest domain classifier
for face recognition classification: development prospects of the face recognition system in the
actual check-in, and the face recognition system in the actual
δ (Y , Fk0 ) = min δ (Y , Fk0 ), Y ∈ ϕ k (12) check-in some problems.
Among them, δ is judged as the distance. If the
0 B. EXPERIMENTAL PROCEDURE
distance between Y and F is smaller than the distance
k (1) Accuracy rate of face recognition system in actual
0
between the two face images in class Fk , then Y check-in
0 The face recognition attendance system using real-time
and Fk can be regarded as a class of pictures, and the video processing is used to count the sign-in rates of students
recognition is successful; otherwise, they are not classified from two colleges and universities, and to compare and
as similar pictures and Y is a non-training sample image. analyze the sign-on accuracy rate of the face recognition
attendance system using real-time video processing.
C. BASIC MODULE OF FACE RECOGNITION (2) The stability of face recognition attendance system
ATTENDANCE SYSTEM BASED ON REAL-TIME VIDEO with real-time video processing
PROCESSING Taking the manual punch card as the control group, the
The complete attendance system consists of a combination face recognition attendance system using real-time video
of multiple modules, each of which performs different processing is used for the check-in statistics, and the number
functions. In order to reduce the complexity of the program
of check-in punch cards is used as the main observation
and facilitate the reuse of codes, it is conducive to maintain
parameter to detect the stability of the face recognition
and manage the entire system. The design of the face
attendance system.
recognition time and attendance system in this system
(3) Analysis of the truancy rate of face recognition
mainly includes several modules of video acquisition
attendance system using real-time video processing
terminal, cable transmission module, data storage, face
Taking the manual punch card as the control group, and
recognition module and computer terminal module [24-25],
using the real-time video processing face recognition
as shown in flow chart2.
attendance system to perform real-time check-in statistics,
observe the students' skip-rate rate of the two methods.
Complete attendance system module
(4) Interface settings of face recognition attendance system
using real-time video processing
Wired Face Computer
By identifying the student's face as the attendance interface,
Video capture
end
transmission Data storage recognition terminal the successful student attendance and time information will
module module module
be displayed on the screen. By selecting face detection and
FIGURE 2. Complete attendance system module recognition, the video images of the attendance students can
be described by the face recognition attendance system
multi-person video.

VOLUME XX, 2020 5

This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/.
 This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI
10.1109/ACCESS.2020.3007205, IEEE Access
Author Name: Face Recognition Attendance System Based on Real-time Video Processing

C. DATABASE DESIGN
This system uses mysql database for data storage. MySQL TABLE I
VIDEO FACE RECOGNITION ACCURACY
has the advantages of fast speed, low cost, small size, open High school A High school B
source, support for multi-threading and multi-processor, Video face recognition
Test Control Test Control
accuracy
good portability, and at the same time can use ODBC (open group group group group
database connection) and SQL (structured query language) Accurate identification
82 94 83 96
(%)
to increase the data (create), delete (delete), query (select),
Video is blurry (%) 15 0 12 0
modify (update) operations. Punch failed (%) 3 6 5 4

D. FACE RECOGNITION MODULE DESIGN B. STABILITY ANALYSIS OF FACE RECOGNITION

This system uses a variety of popular programming ATTENDANCE SYSTEM BASED ON REAL-TIME VIDEO
languages-Python (Interpretive Language), Java, C ++ and PROCESSING
other languages to develop. Python is used for real-time The face recognition attendance system for real-time video
video data collection and face recognition processing. C ++ processing performs video punching every two hours.
is used to complete the related system running components Comparing the data of the face recognition attendance system
such as file operations and client running interface design. and manual fingerprint punching, the sign-in success is 1 and
Java is used to build a face recognition WEB platform the failure is 0. The data collection results are shown in Table
service. Combined with the third-party library Open CV to 2 and Figure 4. An error occurred in the face recognition
realize the writing of face recognition module, OpenCV attendance system from 7 am to 9 am; from 13 noon to 21 pm,
provides rich visual processing and image processing the face recognition attendance system and manual
algorithms to analyze and process the facial features fingerprint punch card can correctly identify the check-in.
captured by the classroom camera. The system is divided The experimental results show that the length of time when
into three modules: front-end operation, real-time video the smart machine is turned on may bring some experimental
face recognition module, and background data errors, indicating that the human-machine interactive testing
management. instrument needs to be pre-powered on for two to four hours
before the accuracy of the testing data can be guaranteed.
V. DISCUSSION
TABLE Ⅱ
A. REAL-TIME VIDEO PROCESSING FACE STABILITY ANALYSIS OF FACE RECOGNITION ATTENDANCE SYSTEM BASED
RECOGNITION ATTENDANCE SYSTEM IN THE ACTUAL ON REAL-TIME VIDEO PROCESSING
SIGN-IN ACCURACY RATE High school A High school B
Time
By investigating two colleges A and B in a province, we Test group Control group Test group Control group
7：00 0 1 1 1
conducted experiments in colleges based on the application
9：00 1 1 0 1
of face recognition attendance system accuracy, and selected 11：00 1 1 1 0
200 college students who need to punch cards. The statistical 13：00 1 1 1 1
results of the survey are shown in Table 1 and Figure 3. 15：00 1 1 1 1
According to the experimental results, the accuracy rate of 17：00 1 1 1 1
face recognition in the classroom video of the two colleges is 19：00 1 1 1 1
currently high, and the accuracy rate of face recognition in 21：00 1 1 1 1
the college video is about 82%. About 15% of the failed card 23：00 1 1 1 1
punches due to video blur and other reasons, About 3% of
the staff failed to punch in, it can be seen from these data that
the accuracy rate of the video face recognition system is
relatively high.

FIGURE 4. Stability analysis of face recognition attendance system

based on real-time video processing

FIGURE 3. Video face recognition accuracy

6 VOLUME XX, 2020

This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/.
 This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI
10.1109/ACCESS.2020.3007205, IEEE Access
Author Name: Face Recognition Attendance System Based on Real-time Video Processing

C. ANALYSIS OF THE SKIPPING RATE OF FACE an error rate of about 20% in the test results. Most students
RECOGNITION ATTENDANCE SYSTEM BASED ON face information errors that will change, such as changes in
REAL-TIME VIDEO PROCESSING facial features, accessories, cosmetics and lighting caused by
Because the face recognition attendance system has a medical plastics, making it impossible to extract the correct
real-time video function, it can count the number of people
logo from the picture; Another main reason is the accuracy of
online in real time. Compared with the traditional
identification code comparison. The system fails to correctly
fingerprint check-in, students' rate of skipping classes will
judge that the two identifiers are very similar. Under various
change. The survey data is shown in Table 3 and Figure 5.
perfect conditions, make sure to adapt to various situations
The results showed that after using the face recognition
.
system, the skipping rate of the two universities was TABLE 4
significantly lower than that of the control group, only TEST NUMBER AND ACCURACY RATE TABLE
about 13%. Therefore, the face recognition attendance Test number Correct rate
system can effectively increase the attendance rate of 10～20 85%
university classrooms, and is an effective method to restrain 20～25 80%
students from skipping classes.
25～30 78%
TABLE Ⅲ 30～35 75%
STATISTICAL TABLE FOR CHECKING STUDENT ATTENDANCE
Check student attendance High school A High school B Correct rate
Not named 93 (93%) 94 (94%) 85%
Traditional roll call 73 (73%) 70 (70%) 80%
78%
Video face recognition 13 (13%) 12 (12%)
75%
Pe
rce
nt
spe age
ed dec
Pe lin line
rce e
nta
tre ge d
nd ow
cu nw
rve ar
d

FIGURE 6. Test number and accuracy rate chart

FIGURE 5. Statistics chart for checking student attendance

V. CONCLUSIONS
D. FACE RECOGNITION ATTENDANCE SYSTEM (1) College attendance management for students has become
INTERFACE SETTINGS one of the hot issues in the society, so the management of
From the amount of change in Table 4 and Figure 6, it can be college students should be strengthened. However, most
seen that by recognizing the student's face as the attendance college students still use traditional manual attendance for
interface, the successful attendance and time information of daily attendance, using paper signatures or teacher orders, but
the confirmed student is displayed on the screen. By selecting now with the gradual rise of technology, some new methods
face detection and recognition, the video images of the point out that gradually, a few colleges and universities will
attendance students can be described by the face recognition use punch card fingerprints and smart attendance methods.
attendance system multi-person video. Compared with the Although there are some ways to stimulate attendance, the
traditional punch card signing, this method greatly improves effect is not so effective. Attendance and these methods have
the efficiency, and can prevent early leave and skip class. By a common shortcoming, fraud will occur, thereby increasing
regularly extracting images, seeing that the students ’handling the rate of absenteeism. This repeated phenomenon not only
of the case is transparent, the inspection method is an has a negative impact on students' psychology and physiology,
imperceptible process to a process where the user perceives but also maintains the normal order of university teaching
no signs, it is a sign of an automated way. After repeated and hinders the quality of teaching. At the same time it will
testing, the test results of the system under different test have a very unfavorable situation training and the formation
sample numbers are as follows: when the input student is 10 of university spirit and discipline.
~ 20, the recognition and sign-in correct rate is 85%; when (2) In this paper, a face recognition attendance system
the input student is 20 ~ 50, the correct rate is 80%. There is based on real-time video processing is designed, and two
VOLUME XX, 2020 7

This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/.
 This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI
10.1109/ACCESS.2020.3007205, IEEE Access
Author Name: Face Recognition Attendance System Based on Real-time Video Processing

colleges in a province are selected for real-time check-in and Cybernetics of Tomsk Polytechnic University., vol. 40, no. 5, pp.
740–745, 2016.
inspection of student attendance. This article mainly sets four [8] Taniya, K., Nidhi, M., Nandini, T. “Automated human resource
directions to consider the problems: the accuracy rate of the and attendance management system based on real time face
face recognition system in the actual check-in, the stability of recognition”, IJSRSET., vol. 16, no. 4, pp. 847-853, 2016.
the face recognition attendance system with real-time video [9] Wu, D., Tang, Y., Q., Lin, G., H., & Hu, H. “Roboust face
recognition based on significance local directional pattern and deep
processing, and the truancy rate of the face recognition learning”, Guangdianzi Jiguang/Journal of Optoelectronics Laser.,
attendance system with real-time video processing It is vol. 27, no. 6, pp. 655-661, 2016.
difficult to analyze the interface settings of the face [10] Sun, Y., Zhao, J., Hu, Y. “Supervised sparsity preserving
recognition attendance system using real-time video projections for face recognition”, Proceedings of Spie., vol. 8009,
no. 4, pp. 357-366, 2017.
processing. By analyzing the situation of these problems, the [11] Best-Rowden, L., & Jain, A., K. “Longitudinal study of automatic
concept of attendance system based on face recognition face recognition”, IEEE Transactions on Pattern Analysis &
technology is proposed, and the research on face recognition Machine Intelligence., vol. PP, no. 99, pp. 1-1, 2018.
attendance system based on real-time video processing is [12] Valentine, T., Lewis, M., B., & Hills, P., J. “Face-space: A unifying
concept in face recognition research”, Quarterly Journal of
carried out. Research data shows that the accuracy of the Experimental Psychology., vol. 69, no. 10, pp. 1996-2019, 2016.
video face recognition system is about 82%. The face [13] Duan, Y., Lu, J., Feng, J., & Zhou, J. “Context-aware local binary
recognition time attendance system and manual fingerprint feature learning for face recognition”, IEEE Transactions on
Pattern Analysis & Machine Intelligence., vol. PP, no. 99, pp. 1-1,
punching are more stable and correctly identify check-ins,
2017.
and the rate of skipping classes is significantly reduced [14] Nemirovskiy, V., B., Stoyanov, A., K., Goremykina, D., S. “Face
compared with the control group, only about 13%. Compared recognition based on the proximity measure clustering”, Institute of
with the control group, the efficiency is greatly improved, Cybernetics of Tomsk Polytechnic University., vol. 40, no. 5, pp.
740–745, 2016.
which can prevent students from leaving early and skipping [15] Chakraborty, S., Singh, S., K., & Chakraborty, P. “Local gradient
classes. hexa pattern: a descriptor for face recognition and retrieval”, IEEE
(3) The attendance system realizes the expected attendance Transactions on Circuits & Systems for Video Technology., vol. PP,
results through face recognition technology with the help of a no. 99, pp. 1-1, 2016.
[16] Best-Rowden, L., & Jain, A., K. “Longitudinal study of automatic
computer, which fully reflects the feasibility design of the face recognition”, IEEE Transactions on Pattern Analysis &
overall algorithm. The students who completed the Machine Intelligence., 2017, vol. 40, no. 1, pp. 148-162, 2018.
attendance sign-in system quickly completed the tasks, got [17] Feng, Q., Yuan, C., Pan, J., S., Yang, J., F., Chou, Y., T., & Zhou,
rid of the complicated sign of roll call, and soon realized the Y. “Superimposed sparse parameter classifiers for face
recognition”, IEEE Transactions on Cybernetics., vol. 47, no. 2, pp.
sign of operation and function. The future system time and 378-390, 2017.
the form of attendance system conversion have made [18] Valentine, T., Lewis, M., B., & Hills, P., J. “Face-space: A unifying
tremendous innovations, greatly improving the attendance concept in face recognition research”, Quarterly Journal of
Experimental Psychology., vol. 69, no. 10, pp. 1996-2019, 2016.
rate and the reliability of face recognition technology. It is
[19] Mudunuri, S., P., & Biswas, S. “Low resolution face recognition
worthy of further exploration and realization by our across variations in pose and illumination”, IEEE Transactions on
scientists. Pattern Analysis & Machine Intelligence., vol. 38, no. 5, pp.
1034-1040, 2016.
REFERENCES [20] Li, Y., Song, W., Cheng, C. “Attendance system of face
recognition based on raspberry pi%”, Microcontrollers and
[1] Solanki, K., Pittalia, P. “Review of face recognition techniques”,
Embedded System Applications., vol. 16, no. 11, pp. 28-30, 34,
International Journal of Computer Applications., vol. 133, no. 12,
2016.
pp. 20-24, 2016.
[21] Bharadwaj, S., Bhatt, H., S., Vatsa, M., & Singh, R. “Domain
[2] Stoll Chloé, Richard, P. G., Roberto, C., Junpeng, L., Dye, M. W.
specific learning for newborn face recognition”, IEEE Transactions
G., & Florent, A. “Face recognition is shaped by the use of sign
on Information Forensics & Security., vol. 11, no. 7, pp. 1-1, 2016.
language”, Journal of Deaf Studies & Deaf Education., vol. 23, no.
[22] Zhu, X., Liu, H., Lei, Z., Shi, H., Yang, F., & Yi, D. “Large-scale
1, pp. 1-9, 2018.
bisample learning on id versus spot face recognition”,
[3] Deng, W., Hu, J., Guo, J. “Face recognition via collaborative
International Journal of Computer Vision., vol. 127, no. 6-7, pp.
representation: its discriminant nature and superposed
684-700, 2019.
representation”, IEEE Transactions on Pattern Analysis and
[23] Nguyen, H., Yang, W., Sheng, B. “Discriminative low-rank
Machine Intelligence., vol. 40, no. 10, pp. 1-1, 2017.
dictionary learning for face recognition”, Neurocomputing., vol.
[4] Pei, T., Zhang, L., Wang, B., Li, F., & Zhang, Z. “Decision
173, no. P3, pp. 541-551, 2016.
pyramid classifier for face recognition under complex variations
[24] Ji, H., K., Sun, Q., S., Ji, Z., X., Yuan, Y., H., & Zhang, G., Q.
using single sample per person”, Pattern Recognition., vol. 64, no.
“Collaborative probabilistic labels for face recognition from single
C, pp. 305-313, 2016.
sample per person”, Pattern Recognition., vol. 62, no. C, pp.
[5] Shi, H., Wang, X., Yi, D., Lei, Z., Zhu, X., & Li, S. Z.
125-134, 2017.
“Cross-modality face recognition via heterogeneous joint bayesian”,
[25] Wang, L., Chen, S. “Joint representation classification for
IEEE Signal Processing Letters., vol. 24, no. 1, pp. 81-85, 2017.
collective face recognition”, Pattern Recognition., vol. 63, no.
[6] Ding, C., Tao, D. “Trunk-branch ensemble convolutional neural
Complete, pp. 182-192, 2017.
networks for video-based face recognition”, IEEE Transactions on
Pattern Analysis & Machine Intelligence., vol. PP, no. 99, pp. 1-1,
2016.
[7] Nemirovskiy, V., B., Stoyanov, A. K., Goremykina, D., S. “Face
recognition based on the proximity measure clustering”, Institute of

8 VOLUME XX, 2020

This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/.
 This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI
10.1109/ACCESS.2020.3007205, IEEE Access
Author Name: Face Recognition Attendance System Based on Real-time Video Processing

HAO YANG was born in Yongkang，

Zhejiang,P.R. China, in 1986. He received
the Doctor degree from University of
Chinese Academy of Sciences, P.R. China.
Now, he works in College of Media
Engineering, Communication University
of Zhejiang. His research interest include
formal method and face recognition.
E-mail: 415360889@qq.com

XIAOFENG HAN was born in Taian, Shandong,

P.R. China, in 1982. He received the master‘s
degree from Shandong University of Science and
Technology, P.R. China. Now, he work in
College of Mathematics and Systems Science,
Shandong University of Science and Technology.
His research interest include image processing,
information and computing science.
E-mail: hs56923@.163.com

VOLUME XX, 2020 9

This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/.